一种带有牛顿方向的自适应扰动改进粒子群优化算法  

An Improved PSO Algorithm with Adaptive Dynamical Disturbance and Newton Direction

作　　者：史旭栋 高岳林[2] 

机构地区：[1]宁夏大学数学与计算机学院,宁夏银川750021 [2]北方民族大学信息与系统科学研究所,宁夏银川750021

出　　处：《兰州文理学院学报（自然科学版）》2017年第6期69-74,共6页Journal of Lanzhou University of Arts and Science(Natural Sciences)

基　　金：国家自然科学基金项目(61561001);北方民族大学重点科研项目(2015KJ10)

摘　　要：针对PSO算法在求解高维复杂优化问题时,往往会陷入早熟,本文引入了随机扰动项,以改进粒子群优化算法的速度更新公式.利用牛顿下降方向引导当前最好粒子快速寻优,从而提出了一种带有牛顿方向的自适应扰动改进粒子群优化算法(ARPSO).该算法能够使被困的粒子跳出局部陷阱,从而实现全局寻优.从理论上证明了该算法的收敛特性,实验结果表明所提算法具有更好的全局寻优能力和更高计算精度.In solving high-dimensional complex optimization problems,the PSO algorithm usually tends to fall into premature.This paper first introduces stochastic disturbance in order to improve the updated formula of particle swarm optimization algorithm,then,by Newton direction,current best particle is guided to find optimal value quickly.Based on the above,an improved PSO algorithm with adaptive dynamical disturbance and Newton direction is proposed,which can make the trapped particles jump out of the local trap to achieve global optimization.The convergence performance of the algorithm is proved theoretically and the experimental results show that the proposed algorithm has better global optimization ability and higher computational accuracy.

关 键 词：粒子群优化 自适应扰动 牛顿方向 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]